3.4 Prevalence of Soil-Pica

The second was: "What is the prevalence rate of soil-pica behavior among children, especially preschool children? among adults? among pregnant women?" The panelists stressed that the prevalence of pica behavior is highly dependent upon how it is defined, and eventually defined the various terms (e.g., soil ingestion, soil-pica, geophagy) used throughout the workshop. Even with a clear definition of soil-pica, however, the panelists had difficulty quantifying the prevalence among subpopulations, given the lack of extensive soil ingestion studies. Nonetheless, the panelists agreed that soil-pica exists and that ATSDR should continue to evaluate the public health implications of all types of soil ingestion behavior, including soil-pica. Following is a summary of the discussion that led to this conclusion.

• The importance of clearly defined terminology. The panelists repeatedly stressed that the prevalence of soil-pica behavior depends on how one defines this behavior. As an example of their concern, panelists noted that they have seen "soil-pica" defined in terms of quantities of soil ingested and whether the ingestion behavior is abnormal, intentional, or repetitive (BL, JM, DV). One panelist cautioned against limiting pica behavior to abnormal soil ingestion levels, because such a restricted definition would overlook the fact that children's typical behaviors can lead to relatively high soil/dust ingestion rates (see Section 3.1) (NF). The other panelists stressed that a clear, unambiguous definition of soil-pica must be crafted so that ATSDR can quantify the prevalence using the various methods discussed in Section 3.3. Given the importance of communicating with consistent terminology, the panelists defined soil ingestion, soil-pica, and geophagy at the close of the meeting. Refer to Section 2.0 for these definitions.

• Comments on the distribution of soil ingestion rates. When discussing the prevalence of soil-pica, two panelists suggested that ATSDR view soil ingestion rates as a continuum, possibly by characterizing the distribution of these rates (SD, JM). Knowing the distribution of soil ingestion rates, according to one panelist, would allow researchers to quantify the distribution of exposures to soil contaminants at sites where the nature and extent of soil contamination has been determined, but this panelist was not convinced that the data currently available are sufficient for estimating this distribution (SD). Another panelist then asked whether a table from a publication (Calabrese and Stanek, 1998) provides a reasonable estimate of the distribution of soil intakes (DM). The panelist responded that he was not sure, because he did not know how the estimate was derived (SD). ⁽²⁾ Given the importance of knowing the distribution of soil ingestion rates for various age groups, one panelist suggested that ATSDR try to estimate these distributions from the various soil ingestion studies that have been published.

• Comments on data reported in the scientific literature. Though not commenting specifically on the prevalence of soil-pica behavior, several panelists noted relevant data documented in the scientific literature. For instance, one panelist noted that a team of researchers has estimated that 33% of children ingest more than 10 grams of soil 1 or 2 days a year (Calabrese and Stanek, 1998) (DM). Another panelist cautioned, however, that this estimate is based on an extrapolation of a short-term study and not on a study of soil ingestion over an entire year (NF). Another panelist noted that his analytical studies as well as those published by Ed Calabrese and Ed Stanek, and by Michael Wong present estimatesAppendix B: Charge to the Panelists of soil ingestion rates, though they do not have consistent findings (SD). Another panelist indicated that his research has found that 30% of children (aged 1-3 years) in Rochester, New York, ingest soils, based on a survey of parents, and that this behavior is associated with a 14% increase in blood lead levels (BL). This panelist cautioned about assuming that these findings might apply to other sites and other contaminants. Overall, the panelists thought their comments confirm that soil-pica exists, but they refrained from providing quantitative estimates of the prevalence of soil-pica behavior, largely because the available studies are limited in duration and not based on a population that represents all groups of children.

• Prevalence of soil-pica among pregnant women. One panelist noted that women in urban areas would likely not dig and process their own geophagical clays, but would likely purchase them or obtain them from areas where they were reared when relatives came to visit (DV). Further, he doubted that pregnant women in urban areas would consume surface soils from their backyards. Consistent with these comments, another panelist provided an anecdotal account of stores in the Atlanta area that sell geophagical clays, which pregnant women might consume (JM). The panelists eventually agreed that studies have not been conducted to determine the extent to which pregnant women exhibit soil-pica behavior, though they suspected that consumption of residential soils is likely rare.

• Variations in soil ingestion rates with age. Citing his own research, one panelist indicated that the percentage of children in his study who ingested soils, as reported by their parents, was 3% for 6-month-old children, 30% for 12-month-old children, 31% for 18-month-old children, and gradually lower percentages for older children (BL). (Note: this research did not ask about "pica behavior," but rather asked parents whether their children ingest soils.) Somewhat consistent with this finding, another panelist added that he believes intentional soil ingestion behaviors decrease as children reach roughly the age of 3 (DV). He attributed this apparent decrease to the observation that parents try to control certain behaviors (including soil ingestion) as children reach ages when they can reason, while they overlook these behaviors when children are younger. One panelist indicated that the available data on this topic are extremely limited (NF).

• Implications of the prevalence of soil-pica behavior. The panelists offered various opinions on the implications of soil-pica behavior, regardless of not knowing the exact extent to which it occurs. One panelist, for example, believed that a significant number of children exhibit soil-pica behavior, but added that an insignificant number of children might develop adverse health effects (NF). She noted that ATSDR ultimately needs to consider many factors other than the prevalence of soil-pica behavior (e.g., the nature and extent of contamination and bioavailability) to put the health concerns into perspective. This panelist added, however, that ATSDR should err on the side of possibly overestimating the prevalence of soil-pica behavior, given that it might be associated with adverse health effects. Another panelist agreed, noting that his research has found soil ingestion, as reported by parents, to be a significant risk factor for childhood lead poisoning (BL). The panelists agreed that ATSDR should continue to evaluate the public health implications of soil-pica behavior, despite the uncertainties associated with the nature and extent of soil-pica. Most panelists agreed that ATSDR should try to validate the public health significance of soil-pica behavior through site-specific studies.

---

^2. At this point, an observer clarified that the author of the publication apparently extrapolated the results of a 2-week analytical study to soil ingestion rates over the course of the year. This extrapolation reportedly assumed that the variability of soil ingestion rates over a year is greater than that which was observed in 2 weeks--an assumption the observer questioned given that the study of concern was conducted during the summer, when soil ingestion rates would likely be greatest and perhaps most variable.

Top of Page